Typically, polymers are either used in a non-covalent fashion, with the drug compound physicochemically formulated into a solvent-polymer mixture, or by permanent covalent attachment of a polymer reagent to one of the drug's functional groups.
Non-covalent drug encapsulation has been applied to depot formulations for long-acting release profiles. Typically, drug is mixed with polymer material and processed in such fashion, that the drug becomes distributed throughout the bulk polymer material. Such polymer-protein aggregates may be shaped as microparticles which are administered as an injectable suspension or they are formulated as gels which are administered in a single bolus injection. Drug release occurs when the polymer swells or degradation of the polymer allows for diffusion of the drug to the exterior. Such degradation processes may be autohydrolytic or enzyme-catalyzed. An example for a marketed drug based on bolus administration of a drug-polymer gel is Lupron Depot. An example for a marketed drug based on suspended microparticles is Nutropin Depot.
A disadvantage of the non-covalent approach is that in order to prevent uncontrolled, burst-type release of the drug, encapsulation has to be highly efficient by creating a sterically highly crowded environment. Restraining the diffusion of an unbound, water soluble drug molecule requires strong van der Waals contacts, frequently mediated through hydrophobic moieties. Many conformationally sensitive therapeutics such as proteins or peptides are rendered dysfunctional during the encapsulation process and/or during subsequent storage. In addition, such amino-containing drug compounds readily undergo side reactions with polymer degradation products (D. H. Lee et al., J. Contr. Rel., 2003, 92, 291-299). Furthermore, dependence of the release mechanism upon biodegradation may cause interpatient variability.
Alternatively, drugs may be conjugated to polymers through permanent covalent bonds. This approach is applied to various classes of molecules, from so-called small molecules, through natural products up to larger proteins.
Many small molecule medicinal agents, like alkaloids and anti-tumor agents, show low solubility in aqueous fluids. One way to solubilize these small molecule compounds is to conjugate them to hydrophilic polymers. A variety of water-soluble polymers, such as human serum albumin, dextran, lectins, poly(ethylene glycol) (PEG), poly(styrene-co-maleic anhydride), poly(N-hydroxypropylmethacrylamide), poly(divinyl ether-co-maleic anhydride), hyaluronic acid have been described for this purpose (R. Duncan, Nature Rev. Drug Disc., 2003, 2, 347-360).
A major challenge in cancer therapy is to selectively target cytotoxic agents to tumor cells. A promising method to accumulate small molecule anticancer agents in tumor tissue and decrease undesirable side effects of these agents is the attachment of the cytotoxin to a macromolecular carrier. The passive targeting of polymeric drug conjugates to tumors is based on the so-called enhanced permeability and retention effect (EPR) as described by Matsumura, Y. and Maeda, H., in Cancer Res., 1986, vol 6, pp 6387-6392. As a result, several polymer-drug conjugates have entered clinical trial as anticancer agents.
Covalent modification of biological molecules with poly(ethylene glycol) has been extensively studied since the late 1970s. So-called PEGylated proteins have shown improved therapeutic efficacy by increasing solubility, reducing immunogenicity, and increasing circulation half-live in vivo due to reduced renal clearance and proteolysis by enzymes (see, for example, Caliceti P., Veronese F. M., Adv. Drug Deliv. Rev. 2003, 55, 1261-1277).
However, many medicinal agents such as INFalfa2, saquinavir or somatostatin are inactive or show decreased biological activity when a polymer is covalently conjugated to the drug molecule (T. Peleg-Shulman et al., J. Med. Chem., 2004, 47, 4897-4904).
In order to avoid shortcomings imposed by either non-covalent polymer mixtures or permanent covalent attachment, it may be preferable to employ a prodrug approach for chemical conjugation of drug to polymer carrier. In such polymeric prodrugs, the biologically active moieties are typically linked to the polymeric carrier moiety by a temporary bond formed between the carrier moiety and a hydroxy, amino or carboxy group of the drug molecule (such as is shown in FIG. 1).
Prodrugs are therapeutic agents that are almost inactive per se but are predictably transformed into active metabolites (see B. Testa, J. M: Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003, page 4). The carrier prodrug approach may be applied in such a fashion that the medicinal agent is released in vivo from the polymer in order to regain its biological activity. The reduced biological activity of the prodrug as compared to the released drug is of advantage if a slow or controlled release of the drug is desired. In this case, a relatively large amount of prodrug may be administered without concomitant side effects and the risk of overdosing. Release of the drug occurs over time, thereby reducing the necessity of repeated and frequent administration of the drug.
Prodrug activation may occur by enzymatic or non-enzymatic cleavage of the temporary bond between the carrier and the drug molecule, or a sequential combination of both, i.e. an enzymatic step followed by a non-enzymatic rearrangement. In an enzyme-free in vitro environment such as an aqueous buffer solution, a temporary bond such as an ester or amide may undergo hydrolysis, but the corresponding rate of hydrolysis may be much too slow and not therapeutically useful. In an in vivo environment, esterases or amidases are typically present and may cause significant catalytic acceleration of the kinetics of hydrolysis from twofold up to several orders of magnitude (see, for example, R. B. Greenwald et al. J. Med. Chem. 1999, 42 (18), 3857-3867).
Definitions Based on IUPAC (as given under http://www.chem.qmul.ac.uk/iupac/medchem/ (accessed on 8 Mar. 2004)
Prodrug
A prodrug is any compound that undergoes biotransformation before exhibiting its pharmacological effects. Prodrugs can thus be viewed as drugs containing specialized non-toxic protective groups used in a transient manner to alter or to eliminate undesirable properties in the parent molecule.
Carrier-Linked Prodrug (Carrier Prodrug)
A carrier-linked prodrug is a prodrug that contains a temporary linkage of a given active substance with a transient carrier group that produces improved physicochemical or pharmacokinetic properties and that can be easily removed in vivo, usually by a hydrolytic cleavage. This is shown graphically in FIG. 1.
Cascade Prodrug
A cascade prodrug is a carrier prodrug for which the cleavage of the carrier group becomes effective only after unmasking an activating group.
Polymeric Cascade Prodrug
A polymeric cascade prodrug is a carrier prodrug that contains a temporary linkage of a given active substance with a transient polymeric carrier group for which the cleavage of the carrier becomes effective only after unmasking an activating group.
Bioprecursor Prodrug
A bioprecursor prodrug is a prodrug that does not imply the linkage to a carrier group, but results from a molecular modification of the active principle itself. This modification generates a new compound, able to be transformed metabolically or chemically, the resulting compound being the active principle.
Biotransformation
Biotransformation is the chemical conversion of substances by living organisms or enzyme preparations.
Prodrugs fall in two classes, bioprecursors and carrier-linked prodrugs. Bioprecursors do not contain a carrier group and are activated by the metabolic creation of a functional group. In carrier-linked prodrugs the active substance is linked to a carrier moiety by a temporary linkage. This invention is concerned with polymeric carrier-linked or macromolecular prodrugs, where the carrier itself is a macromolecule such as a carrier protein or polysaccharide or polyethylene glycol. Specifically, the invention relates to polymeric carrier-linked prodrugs for which this cleavage between polymer and drug proceeds in two steps according to a cascade mechanism.
Cleavage of a carrier prodrug generates a molecular entity (drug) of increased bioactivity and at least one side product, the carrier. This side product may be biologically inert (for instance PEG) or may have targeting properties (for instance antibodies). After cleavage, the bioactive entity will reveal at least one previously conjugated and thereby protected functional group, and the presence of this group typically contributes to the drug's bioactivity.
In order to implement a prodrug strategy, at least one certain functional group in the drug molecule is employed for attachment of the carrier polymer. Preferred functional groups are hydroxyl or amino groups. Consequently, both the attachment chemistry and hydrolysis conditions vary greatly between these two functionalities.
In a simple one-step mechanism, the prodrug's temporary linkage is characterized by an intrinsic liability or enzyme dependence. The susceptibility of this linkage to hydrolysis in an aqueous environment with our without enzyme catalysis controls the cleavage kinetics between polymeric carrier and drug. Numerous macromolecular prodrugs are described in the literature where the temporary linkage is a liabile ester bond. In theses cases, the functional group provided by the bioactive entity is either a hydroxyl group or a carboxylic acid (e.g. Y. Luo, M R Ziebell, G D Prestwich, “A Hyaluronic Acid—Taxol Antitumor Bioconjugate Targeted to Cancer Cells”, Biomacromolecules 2000, 1, 208-218, J Cheng et al, Synthesis of Linear, beta-Cyclodextrin Based Polymers and Their Camptothecin Conjugates, Bioconjugate Chem. 2003, 14, 1007-1017, R. Bhatt et al, Synthesis and in Vivo Antitumor Activity of Poly(L-glutamic acid) Conjugates of 20(S)-Camptothecin, J. Med. Chem. 2003, 46, 190-193; R. B. Greenwald, A. Pendri, C. D. Conover, H. Zhao, Y. H. Choe, A. Martinez, K. Shum, S. Guan, J. Med. Chem., 1999, 42, 3657-3667; B. Testa, J. M: Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003, Chapter 8).
Especially for therapeutic biomacromolecules but also for certain small molecule drugs, it may be desirable to link the macromolecular carrier to amino groups of the bioactive entity (i.e. N-terminus or lysine amino groups of proteins). This will be the case if masking the drug's bioactivity requires conjugation of a certain amino group of the bioactive entity, for instance an amino group located in an active center or a region or epitope involved in receptor binding. Also, during preparation of the prodrug, amino groups may be more chemoselectively addressed and serve as a better handle for conjugating carrier and drug because of their greater nucleophilicity as compared to hydroxylic or phenolic groups. This is particularly true for proteins which may contain a great variety of different reactive functionalities, where non-selective conjugation reactions lead to undesired product mixtures which require extensive characterization or purification and may decrease reaction yield and therapeutic efficiency of the product.
Amide bonds as well as aliphatic carbamates are much more stable towards hydrolysis than ester bonds, and the rate of cleavage would be too slow for therapeutic utility in a carrier-linked prodrug. Therefore it is advantageous to add structural chemical components such as neighbouring groups in order to exert control over the cleavability of the prodrug amide bond. Such additional cleavage-controlling chemical structures that are not provided by the carrier entity nor by the drug are called linker. Prodrug linkers can have a strong effect on the rate of hydrolysis of a given temporary bond. Variation of the chemical nature of these linkers allows to engineer the linker properties to a great extent.
For instance, prodrug linkers may be designed for enzyme-selectivity. Prerequisite for enzymatic dependence is that the linker structure displays a structural motif that is recognized as a substrate by a corresponding endogenous enzyme (FIG. 2).
Enzyme-catalyzed acceleration of prodrug cleavage is a desirable feature for organ or cellular targeting applications. Targeted release of the bioactive entity is effected, if an enzyme, that selectively cleaves the linkage, is specifically present in the organ or cell-type chosen for treatment.
A typical property of an enzyme-dependent temporary linkage is its stability with respect to hydrolysis. The temporary linkage itself will not undergo autohydrolysis at a rate that would release drug to such an extent that a therapeutic effect could be induced in a normal dosing regime. It is only in the presence of the enzyme, that the attack of the enzyme on the linkage causes a significant acceleration of cleavage and concomitant an enhancement of free drug concentration.
Several examples have been published for the prodrug activation of amine-containing biologically active moieties by specific enzymes for targeted release. In these cases, cleavage occurs in a one-step process which is catalyzed by the enzyme. G. Cavallaro et al., Bioconjugate Chem. 2001, 12, 143-151 describe the enzymatic release of an antitumoral agent by the protease plasmin. Cytarabin is coupled via the tripeptide sequence D-Val-Leu-Lys to the polymer alpha, beta-poly(N-hydroxyethyl)-DL-aspartamide (PHEA). Enzymatic release of cytarabin is effected by the protease plasmin which concentration is relatively high in various kinds of tumor mass.
Further examples for antitumoral polymeric prodrugs activated by specific enzymes like beta lactamase (R. Satchi-Fainaro et al., Bioconjugate Chem. 2003, 14, 797-804) and cysteine proteases like cathepsin B (R. Duncan et al. J. Contr. Release 2001, 74, 135-146) have been described. Wiwattanapatapee et al. (2003) outline a dendrimer prodrug for colonic delivery of 5-aminosalicylic acid. The drug molecule is conjugated by an azo bond to “generation 3” PAMAM dendrimer. 5-aminosalicylic acid is released in the colon by a bacterial enzyme called azo reductase (W. R. Wiwattanapatapee, L. Lomlim, K. Saramunee, J. Controlled Release, 2003, 88: 1-9).
A. J. Garman et al. (A. J. Garman, S. B. Kalindjan, FEBS Lett. 1987, 223 (2), 361-365 1987) use PEG5000-maleic anhydride for the reversible modification of amino groups in tissue-type plasminogen activator and urokinase. Regeneration of functional enzyme from PEG-uPA conjugate upon incubation at pH 7.4 buffer by cleavage of the maleamic acid linkage follows first order kinetics with a half-life of 6.1 h. The prodrug cleavage was not investigated in the presence of enzymes, and it can be expected—as explained above—that proteases present in the in vivo environment will significantly contribute to the cleavage of the temporary amide linkage. A further disadvantage of this linkage is the lack of stability of the conjugate at lower pH values. This limits the applicability of the linker to active agents which are stable at basic pH values, as purification of the active agent polymer conjugate has to be performed under basic conditions to prevent premature prodrug cleavage.
Cascade mechanisms have proven particularly useful in the controlled release of drugs containing amino-group functionalities because linker cleavage characteristics can be optimized with greater flexibility than in simple one-step prodrugs.
Cascade cleavage is enabled by linker compounds that are composed of a structural combination of a masking group and an activating group. The masking group is attached to the activating group by means of a first temporary linkage such as an ester or a carbamate. The activating group is attached to an amino-group of the drug molecule through a second temporary linkage, for instance a carbamate. The stability, or susceptibility to hydrolysis of the second temporary linkage is dependent on the presence or absence of the masking group. In the presence of the masking group, the second temporary linkage is highly stable and unlikely to release drug with therapeutically useful kinetics. In the absence of the masking group, this linkage becomes highly liabile, causing rapid cleavage and drug release.
Cleavage of the first temporary linkage is the rate-limiting step in the cascade mechanism. This first step may induce a molecular rearrangement of the activating group such as a 1,6-elimination. The rearrangement renders the second temporary linkage so much more liabile that its cleavage is induced. Ideally, the cleavage rate of the first temporary linkage is identical to the desired release rate for the drug molecule in a given therapeutic scenario. Furthermore, it is desirable that cleavage of the second temporary linkage is instantaneous after its liability has been induced by cleavage of the first temporary bond.
A variety of examples exist for cascade carrier prodrugs where the masking group functionality is performed by the carrier polymer itself as shown diagrammatically in FIG. 3. In the systems discussed below, the masking group is not only part of the carrier but has also been engineered for enzyme-dependence (FIG. 4). Only in the presence of a corresponding enzyme is the rate of cleavage of the first temporary linkage sufficiently accelerated for therapeutic use.
R. B. Greenwald, A. Pendri, C. D. Conover, H. Zhao, Y. H. Choe, A. Martinez, K. Shum, S. Guan, J. Med. Chem., 1999, 42, 3657-3667 & PCT Patent Application WO-A-99/30727 describe a methodology for synthesizing poly(ethylene glycol) prodrugs of amino-containing small molecule compounds based on 1,4- or 1,6-benzyl elimination. In this approach, poly(ethylene glycol) as the polymeric carrier is attached to the benzyl group by means of a first temporary linkage such as an ester, carbonate, carbamate, or amide bond. The benzyl group serves as the activating group, and the PEG polymer also has the function of the masking group in this cascade cleavage mechanism. The amino group of the drug molecule is linked via a second temporary linkage, containing a carbamate group, to the benzyl moiety. The release of PEG from the drug molecule is initiated by enzymatic cleavage of the first temporary linkage followed by a rapid 1,4- or 1,6-benzyl elimination, initiating cleavage of the second temporary linkage.
The same linker system is also used for releasable poly(ethylene glycol) conjugates of proteins (S. Lee, R. B. Greenwald et al. Bioconj. Chem. 2001, 12 (2), 163-169). Lysozyme is used as model protein because it loses its activity when PEGylation takes place on the epsilon-amino group of lysine residues. Various amounts of PEG linker were conjugated to the protein. Regeneration of native protein from the PEG conjugates occurrs by enzymatic cleavage in rat plasma or in non-physiological high pH buffer.
Greenwald et al. published in 2000 a poly(ethylene glycol) drug delivery system of amino-containing prodrugs based on trimethyl lock lactonization (R. B. Greenwald et al. J. Med. Chem. 2000, 43(3), 457-487; PCT Patent Application No. WO-A-02/089789). In this prodrug system, substituted o-hydroxyphenyl-dimethylpropionic acid is linked to PEG by an ester, carbonate, or carbamate group as a first temporary linkage and to amino groups of drug molecules by means of an amide bond as second temporary linkage. The rate-determining step in drug release is the enzymatic cleavage of the first linkage. This step is followed by fast amide cleavage by lactonization, liberating a potentially toxic aromatic lactone side product.
Similar prodrug systems were described by F. M. H. DeGroot et al. (WO02083180 and WO04043493A1) and D. Shabat et al. (WO04019993A1). WO02083180 discloses a prodrug system with elongated and multiple linkers based on 1,(4+2n) elimination. The masking moieties in these examples were specifically designed for enzymatic cleavage. This approach was extended to dendritic prodrug system where one enzymatic activating event triggered the release of more than one drug molecule (WO04043493A1). WO04019993A1 discloses a similar prodrug system based on a self-immolative dendrimer releasing many drug moieties upon a single enzymatic activating event. These systems are characterized by the absence of a polymeric carrier. Instead, oligomerization of prodrug linker components provides for a high molecular weight of the prodrug, and prodrug cleavage generates linker residues and free drug, but no polymeric entity is released.
The disadvantage in the abovementioned prodrug systems described by Greenwald, DeGroot and Shabat is the release of potentially toxic aromatic small molecule side products like quinone methides after cleavage of the temporary linkage. The potentially toxic entities are released in a 1:1 stoichiometry with the drug and can assume high in vivo concentrations. This risk factor is even greater if self-immolative dendritic structures based on oligomers of the activating group are employed and more aromatic side products than drug molecules are released.
More recently, R. B. Greenwald et al. (Greenwald et al. J. Med. Chem. 2004, 47, 726-734) described a PEG prodrug system based on bis-(N-2-hydroxyethyl)glycin amide (bicin amide) linker. In this system two PEG molecules are linked to a bicin molecule coupled to an amino group of the drug molecule. The first two steps in prodrug activation is the enzymatic cleavage of both PEG molecules. Different linkages between PEG and bicin are described resulting in different prodrug activation kinetics. The main disadvantage of this system is the slow hydrolysis rate of bicin amide conjugated to the drug molecule (t1/2=3 h in phosphate buffer) which results in the release of a bicin-modified prodrug intermediate that may show different pharmacokinetic and pharmacodynamic properties as compared to the parent drug molecule.
Cascade prodrugs with masking groups that are part of the carrier polymer are limited in the control of drug release kinetics. As masking group cleavage is the rate-limiting step in the cascade mechanism, its molecular structure governs the kinetics. If the carrier polymer is identical to the masking group, the structural flexibility is restricted to the polymers' features. Alternatively, if the polymer requires structural modification in order to match the requirements for controlled cleavage, synthesis of corresponding structures may become more difficult. Also, the incorporation of masking group features into a polymer may change its safety profile.
Therefore is it preferred to structurally separate the masking group and the carrier. This may be achieved by employing a permanent bond between polymer carrier and activating group. This stable bond does not participate in the cascade cleavage mechanism. If the carrier is not serving as a masking group and the activating group is coupled to the carrier by means of a stable bond, release of potentially toxic side products such as the activating group is avoided. The stable attachment of activating group and polymer also suppresses the release of drug-linker intermediates with undefined pharmacology.
Systems have been developed for targeted delivery of therapeutic agents by rendering the masking group enzyme-dependent. Only in the presence of a corresponding enzyme is the rate of cleavage of the first temporary linkage connecting the masking group with the activating group sufficiently accelerated for therapeutic use.
Antczak et al. (Bioorg Med Chem 9 (2001) 2843-48) describe a reagent which forms the basis for a macromolecular cascade prodrug system for amine-containing drug molecules. In this approach an antibody serves as carrier, a stable bond connects the antibody to an activating moiety, carrying an enzymatically cleavable masking group. Upon enzymatic removal of the ester-linked masking group, a second temporary bond cleaves and releases the drug compound, as shown in FIG. 6.
D. Shabat et al. (Chem. Eur. J. 2004, 10, 2626-2634) describe a polymeric prodrug system based on a mandelic acid activating moiety. In this system the masking group is linked to the activating moiety by a carbamate bond. The activating moiety is conjugated permanently to a polyacrylamide polymer via an amide bond. After enzymatic activation of the masking group by a catalytic antibody, the masking group is cleaved by cyclization and the drug is released. The activating moiety is still connected to the polyacrylamide polymer after drug release.
M.-R. Lee et al. describe (Angew. Chem. 2004, 116, 1707-1710) a similar prodrug system based on mandelic acid activating moiety and an enzymatically cleavable ester-linked masking group.
In all of these described prodrug-polymer systems the masking group is specifically designed to be substrate to an enzyme, and masking group cleavage will almost entirely depend upon enzymatic catalysis with the disadvantages of interpatient variability, injection site variability and poor in vitro-in vivo correlation.
A major drawback of predominantly enzymatic cleavage is interpatient variability. Enzyme levels may differ significantly between individuals resulting in biological variation of prodrug activation by enzymatic cleavage. Enyzme levels may also vary depending on the site of administration, for instance it is known that in the case of subcutaneous injection, certain areas of the body yield more predictable therapeutic effects than others. To reduce this unpredictable effect, non-enzymatic cleavage or intramolecular catalysis is of particular interest (see, for example, B. Testa, J. M: Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003, page 5).
Furthermore, it is difficult to establish an in vivo-in vitro correlation of the pharmacokinetic properties for such enzyme-dependent carrier-linked prodrugs. In the absence of a sound in vivo-in intro correlation the optimization a release profile becomes a cumbersome task.
Also, the need for enzyme selectivity imposes a severe limitation on the structural features that can be used in the prodrug linker. This restriction greatly hinders the development of a sound structure-activity relationship and consequently the optimization of linker cleavage kinetics.
For these reasons, there is a need to provide novel linker and/or carrier technologies for forming polymeric prodrugs of amine containing active agents in order to overcome the limitations of the described polymeric prodrugs.